You are viewing a plain text version of this content. The canonical link for it is here.
Posted to notifications@iotdb.apache.org by "Jialin Qiao (Jira)" <ji...@apache.org> on 2022/07/19 09:48:00 UTC

[jira] [Assigned] (IOTDB-3887) [ shrink + query ] The distributed plan of the query is sent to the node that is executing the region data migration

     [ https://issues.apache.org/jira/browse/IOTDB-3887?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Jialin Qiao reassigned IOTDB-3887:
----------------------------------

    Assignee: Gaofei Cao

> [ shrink + query ] The distributed plan of the query is sent to the node that is executing the region data migration
> --------------------------------------------------------------------------------------------------------------------
>
>                 Key: IOTDB-3887
>                 URL: https://issues.apache.org/jira/browse/IOTDB-3887
>             Project: Apache IoTDB
>          Issue Type: Bug
>          Components: mpp-cluster
>            Reporter: 刘珍
>            Assignee: Gaofei Cao
>            Priority: Major
>         Attachments: image-2022-07-19-15-39-54-354.png, ip3_config.properties, ip4_config.properties, screenshot-1.png, screenshot-2.png, screenshot-3.png
>
>
> 3副本3C7D ,缩容1个datanode,缩容成功,
> 执行查询,分布式计划发送到正在执行region数据迁移的节点上,导致查询结果是过期数据。
> region数据迁移成功前,不应该提供查询服务?
> 复现流程
> 1. 私有云环境,准备测试环境
> 准备环境(8C32G机器):
> master_0718_967cde6
> ip2/3/4/5/13/14/16  共7个datanode
> ip3/4/5 共3个confignode
> ratis协议,3副本3C3D
> ip2启动3bm分别连3datanode
> schema_region_consensus_protocol_class=org.apache.iotdb.consensus.ratis.RatisConsensus
> data_region_consensus_protocol_class=org.apache.iotdb.consensus.multileader.MultiLeaderConsensus
> data_replication_factor=3
> schema_replication_factor=3
> max_waiting_time_when_insert_blocked=3600000
> query_timeout_threshold=3600000
> MAX_HEAP_SIZE="16G"
> benchmark测试机:
> 172.20.70.15
> bm_0620_7ec96c1
> 2.  启动ConfigNode
> 启动DataNode
> 启动2个benchmark,分别连ip3,ip4,配置文件见附件。
> 全部是写入顺序数据。
> 缩容前的region信息
>   !screenshot-3.png! 
> 3. benchmark运行3小时
> 4. 缩容ip3,成功
> 缩容后的region信息
>  !screenshot-1.png! 
> 5.  连ip5执行查询
> select count(s_1) from root.ip3.g_0.d_38;
> 多执行几次,就有分布式计划发送到region数据迁移新节点ip4
>  !screenshot-2.png! 
> 分布式查询计划(ip5的log  log-all-20220719.34.log):
>  !image-2022-07-19-15-39-54-354.png! 



--
This message was sent by Atlassian Jira
(v8.20.10#820010)